volume 2 number i o October 1975

Nucleic Acids Research

Replication of double-stranded RNA in particles of Penicillium stoloniferum virus S

K.W.Buck Department of Biochemistry, Imperial College of Science and Technology, London SW7 2AY, U.K. Received 28 August 1975 ABSTRACT RNA polymerase activity was assayed in different particle classes of Penicillium stoloniferum virus S. RNA polymerase activity was found to be associated with H particles, which contain double-stranded RNA and singlestranded RNA, but not with L particles, which contain only double-stranded RNA and not with M particles, which contain only single-stranded RNA. In H particles the reaction occurred with the formation of one new molecule of double-stranded RNA (or two complementary single strands of RNA) per virus particle and the production of product particles (P particles), which contained two molecules of double-stranded RNA (or its equivalent). This RNA polymerase is therefore a replicase, which catalyses the synthesis of the two complementary strands of double-stranded RNA in a single virus particle. This is the first report of this type of RNA polymerase system. INTRODUCTION RNA polymerase activity has been shown to be associated with virions (or cores derived from virions by proteinase action) of double-stranded RNA viruses of animals and higher plants, such as reovirus, cytoplasmic 1 polyhedrosis virus and wound tumour virus . In these cases the enzymes function as transcriptases and the multiple copies of virus messenger RNA molecules produced are released from the particles.

In addition sub-viral

particles, isolated from reovirus infected cells, were found to contain a different RNA polymerase activity, synthesising double-stranded RNA on a single-stranded RNA template; in this case the RNA product remained associated with the sub-viral particles, which are believed to be precursors 2 of mature virions . Recently RNA polymerase activity has been detected in particles of double-stranded RNA viruses from a number of fungi (Penicillium

3 stoloniferum , Penicillium chrysogenum

4

5 and Aspergillus foetidus ) . In the

case of P. stoloniferum viruses the product of reaction was double-stranded RNA of the same molecular weight as virus genomic RNA, and remained within 6 the particles . Because of the limited amount of RNA synthesis obtained, 6 it was postulated by Chater & Morgan as a working model that the synthesis

1889

Nucleic Acids Research consisted of "filling-in" of short single-stranded RNA t a i l s on predominantly double-stranded RNA molecules.

In the present paper i t i s shown that, in

the case of P. stoloniferum virus S, the virus particle RNA polymerase i s a replicase, which, using virus genome double-stranded RNA as template, catalyses the formation of one new double-stranded RNA molecule per virus particle. METHODS

Preparation and purification of virus and virus RNA.

Penicillium

stoloniferum, strain ATCC 14586, was grown in shaken flasks for 2 days and virus was obtained from the homogenised mycelium as described previously , except that the stage involving co—precipitation with yeast RNA was omitted.

8 Virus preparations were purified by sucrose density gradient centrifugation and separation of P. stoloniferum virus S and P. stoloniferum virus F, was carried out as described by Buck & Kempson-Jones . Virus preparations were finally dialysed against 30 mM tris-HCl buffer, pH 8.0, sterilised by f i l t r a t i o n through Millipore GSWPO1300 f i l t e r s and stored at 4 C. Virus RNA was prepared by phenol extraction as described previously . RNA polymerase assays. The RNA polymerase reaction mixture contained 90 raM tris-HCl buffer, pH 8.0; 20 mM magnesium chloride; 1.0 mM EDTA; 0.30 mM ATP; 0.30 mM GTP; 0.30 cM CTP; 0.3OroMH-UTP (51 mCi/ramole). Reaction was started by mixing equal volumes of RNA polymerase reaction mixture and virus preparation (E -^ 0.1 to 1.5) at 30°C. Measurement of incorporation of

P into acid insoluble RNA was carried out as described by Chater H-UTP 6

& Morgan

Isopycnic centrifugation in sucrose density gradients.

Virus preparations

(l ml, E r^ 8) were layered onto pre-fonned linear gradients of 25 to 60% W

/w sucrose in 30 mM tris-HCl buffer, pH 8.0 (60 ml) and centrifuged at

24,000 rev/min in a Beckman SW 25.1 rotor for 18 h at 4 C.

Fractions

(0.5 ml) were collected using an ISCO Model D density gradient fractionator and U.V. analyser.

Individual fractions containing virus particles were

dialysed against 30 mM tris-HCl buffer, pH 8.0 ( 4 x 5

!)•

Isopycnic centrifugation in caesium chloride density gradients. a) Analytical.

Solid caesium chloride was added to virus preparations in

order to give a density of I.36 g/ml.

The solutions were then centrifuged

to equilibrium (18 to 24 h) at 40,000 rev/min at 22 to 24 C using cells with double-sector filled Epon centrepieces in the AN-F rotor in a Beckman Model E analytical ultracentrifuge.

Profiles were obtained with the U.V.

Scanner at 265 nra. Average densities of caesium chloride solution were

1890

Nucleic Acids Research calculated from refractive indices, measured with a Bellinghara and Stanley Abbe High Accuracy '60' refractometer . Virus buoyant densities were calculated by the method of Erikson & Szybalski b) Preparative.

Virus preparations (lml, E ^

. 8) were layered onto pre-

formed linear gradients of 10 to 45% /w caesium chloride in 30 nW trisHC1 buffer, pH 8.0 (60 ml) and centrifuged for 18 h at 24,000 rev/min at 4 C in a Beckman SV 25.1 rotor.

Fractions (0.5 ml) were collected using

an ISCO Model D gradient fractionater and U.V. analyser.

Densities of

g

caesium chloride solutions were calculated from refractive indices . For isolating P particles from RNA polymerase reactions of virus fractions isolated from isopycnic sucrose density gradients, samples were dialysed against 30 mM tris-HCl buffer, pH 8.0 (5 x 5 1) and then adjusted to a density of I.36 by addition of solid caesium chloride.

The solutions

were centrifuged at 45,000 rev/min at 4 C in a Beckman SW 50 rotor for 18 h.

Fractions (0.05 ml) were collected from the bottom of gradients

using a Buchler density gradient fractionator, and diluted to 0.3 nil prior to measurement of extinction or radioactivity. For assay of

H in fractions from caesium chloride density gradients,

samples (0.05 ml) were mixed with 2-methoxyethanol (5 ml) and scintillant (10 ml, 0.6% butyl PBD + 5% naphthalene in toluene), prior to scintillation counting. Measurement of specific activity of RNA in P particles.

P particles, from

RNA polymerase reactions of virus fractions isolated from sucrose density gradients, were isolated by isopycnic caesium chloride density gradient centrifugation and dialysed against 1.5 mM sodium citrate buffer, pH 7.0, containing 15 nM NaCl (5 X 5 ! ) •

Samples (0.5 ml) were made 0.2% with

respect to sodium dodecyl sulphate and incubated at 60 C for 20 min in order to disrupt the virus particles and release the virus RNA Samples were then lyophilised and taken up in water (0.05 ml).

After

heating to 60 C for 10 min and cooling, samples were analysed by electrophoresis in 4% polyacrylamide gels for 2 h as described by Buck & Kerapson-Jones , except that the gels were prepared and run in silica tubes.

After electrophoresis gels were scanned at 250 nm using a Gilford

Model 240 spectrophototneter equipped with a Model 2-410-S Linear Transport. The amount of RNA in the gels was quantitated* from the areas of the peaks, using gel scans obtained from solutions of P. stoloniferum virus S L particle RNA for calibration. After removal from the silica tubes the gels were sliced into 1 mm segments using a Mickle gel slicer and the gel slices were digested and

1891

Nucleic Acids Research counted for

H r a d i o a c t i v i t y as described by Chater & Morgan .

In

order t o monitor the e f f i c i e n c y of counting, samples of the RNA polymerase r e a c t i o n mixture (without virus) were diluted 1 to 40 with 1 mM tris-HCl b u f f e r , pH 8 . 0 , and 0.05 ml samples were subjected to polyacrylaraide gel e l e c t r o p h o r e s i s as above for 1 h.

After t h i s time the

t r a v e l l e d about 4 cm down the gel. and counted for

H-UTP had

The g e l s were then s l i c e d ,

digested

H r a d i o a c t i v i t y under the same conditions as for the

RNA samples.

RESULTS Location of RNA polymerase activity in H particles. P. stoloniferum virus S has been shown previously

to consist of

a number of particle types, readily separated by isopycnic centrifugation in caesium chloride density gradients, namely, LI and L2 particles, each containing one molecule of double-stranded RNA of molecular weight 0.94 x 10

(RNA - Si) and 1.11 x 10

(RNA - S2) respectively, Ml and M 2

particles, each containing one molecule of single-stranded RNA of molecular weight 0.47 x 10 and O.56 x 10 respectively and a range of H particles, containing one molecule of double-stranded RNA (either RNA-S1 or RNA-S2) together with different amounts of single-stranded RNA, which range in sedimentation coefficient from 0 to 15 S.

It was postulated

that H particles were in the process of replicating their RNA; the different lengths of single-stranded RNA would represent different stages of the replication process.

It was therefore important to establish in

which of the component particles of this virus the RNA polymerase activity lay.

A virus preparation was therefore centrifuged to equilibrium in a

preformed linear gradient of sucrose and the gradient was then fractionated. The resultant separation is shown in Fig. 1. Each fraction was dialysed against 30 mM tris-HCl buffer, pH 8.0, and incubated with the RNA polymerase reaction mixture until the reaction was complete (18 h ) .

H-

UTP incorporation into acid insoluble RNA was measured for each fraction. The results (Fig. l) show that there was no •'H-UTP incorporation by M particles, but that the RNA polymerase activity lies in the L to H region of the gradient. In order to ascertain if RNA synthesis takes place in both L and H particles and to establish whether or not RNA synthesis was taking place by filling in of short single-stranded tails on predominantly doublestranded RNA molecules as suggested by Chater & Morgan , use was made of the observation that the product of RNA polymerase action remains within

1892

Nucleic Acids Research

2 Fig. 1.

4

10 12 14 16 id FRACTION NUMBER

6

H-UTP incorporation by fractions of P. stoloniferum virus S obtained by isopycnic centrifugation in a sucrose density gradient.

the particles.

The buoyant density of the particles in caesium chloride

gradient centrifugation should therefore be increased following RNA synthesis.

If the reaction consists of merely filling in of single-

stranded RNA tails, there would be very little increase in buoyant density.

On the other hand if the reaction consisted of synthesising

new molecules of double-stranded RNA the increase in buoyant density would be considerable.

A virus preparation was therefore incubated

with the RNA polymerase reaction mixture and the reaction was allowed to proceed to completion (l8 h ) .

The reaction mixture was dialysed against

30 mM tris-HCl buffer, pH 8.0, and solid caesium chloride was added to give an average density of 1.36.

One portion of the solution was

centrifuged to equilibrium at 45,000 rev/min in the Beckman SW 50 rotor and then fractionated.

The extinction at 260 nm and

H-incorporation

of each fraction are shown in Fig. 2. The radioactivity lay entirely in the product particles, designated P.

M particles were unchanged in

density or amount confirming that these have no polymerase activity. L particles were also unchanged in density or amount, showing that these particles were unreactive and, since there was no

H-incorporation

in this fraction, eliminating the possibility of filling in of singlestranded RNA tails for these particles. In order to obtain higher resolution a second portion of the

1893

Nucleic Acids Research

K0 -1-35 -V30

X

I o o 10 Fig.

2.

15 20 FRACTION NUMBER

Fractionation of the products of P. stoloniferum virus S RNA pDlymerase reaction by isopycnic centrifugation in a caesium chloride density gradient.

solution was centrifuged to equilibrium in the Beckman Model E analytical ultracentrifuge and the u.v. scan, together with a similar profile of unreacted virus, is shown in Fig. 3(a and b ) .

It is clear that RNA

synthesis occurred in the whole range of H particles which were converted to the more dense P particles; two classes of P particles, PI and P2, were resolved with densities of 1-387 and 1.393 respectively.

These

profiles also confirm that M and L particles were unchanged in density or amount. Inclusion of S-adenosyl methionine (SAM) in the RNA polymerase reaction mixture at final concentrations of 1 uM, 10 uM or 100 uM had no effect on the rate or final amount of RNA synthesis achieved by P. stoloniferum virus S.

Moreover when the products, in reactions containing

1 uM, 10 pM or 100 uM SAM, were analysed by preparative and analytical caesium chloride gradient centrifugation, they were found to be identical to those in which reactions were carried out in the absence of SAM.

Profiles

identical to those shown in Fig. 2 and Fig. 3(b) were obtained and all the H-UTP incorporation was found in P particles-

M and L particles were

unchanged in density or amount, while the whole range of H particles was converted to P particles. Nature and location of newly synthesised RNA When RNA was prepared from P particles, isolated by caesium chloride

1894

Nucleic Acids Research

H—h(c) o

Id)

1-332 Fig.

3-

1-358 1-387 1-393 Density of CsCI (g/ml)

Analytical equilibrium caesium chloride density gradient centrifugation of P. stoloniferum virus preparations before and after RNA polyraerase reaction: U.V. scanner tracings, (a) and (b) unfractionated virus before and after RNA polymerase reaction; (c) and (d) virus fraction 11 from isopycnic sucrose gradient centrifugation (Fig. l) before and after RMA polymerase reaction.

density gradient centrifugation, and examined by polyacrylamide gel electrophoresis, only two bands of RNA with mobilities of virus genomic double-stranded RNA i.e. Sl-RNA and S2-RNA were obtained; the slow moving band of single-stranded RNA found after electrophoresis of RNA from H particles

was not detected.

After treatment of P particle RNA with

pancreatic ribonuclease (200 ng/ml, 2 h, 25 C) in 15 nM sodium citrate buffer, pH 7.0, containing 150 mM NaCl, followed by polyacrylamide gel electrophoresis, the same two RNA bands were unchanged in amount or mobility. for

When the gels were fractionated and the gel slices were counted

H-incorporation, all the radioactivity was found in these two RNA

bands.

Similar results were obtained with RNA prepared from reactions

1895

Nucleic Acids Research containing 1 uM, 10 uM or 100 uM SAM. The experiments confirm the results 6 of Chater & Morgan using RNA prepared from reacted, but unfractionated virus, that the only product of reaction is double-stranded RNA with the same molecular weight as virus genome RNA.

When RNA was prepared from

separated Pi and P2 particles, and examined by polyacrylamide electrophoresis it was found that Pi particles gave rise only to Sl-RNA, while P2 particles gave rise only to S2-RNA. Amounts of newly synthesised RNA in P particles In order to quantitate the amount of RNA synthesis achieved by different density classes of H particles, fractions from isopycnic sucrose density gradient centrifugation of the virus (Fig. l) were dialysed against 30 mM tris-HCl buffer, pH 8.0, and incubated with the RNA polymerase reaction mixture until the reaction was complete (l8 h ) .

After dialysis

against the tris buffer to remove exess nucleotides, each fraction was analysed by analytical density centrifugation.

and preparative caesium chloride equilibrium Fractions were also analysed by caesium chloride

analytical ultracentrifugation before polymerase reaction for comparison. Befor reaction, fraction 10 was found to contain only Ll particles and no detectable H particles. No RNA polymerase activity was detected in this fraction, confirming that the particles containing only doublestranded RNA were inactive in the RNA polymerase assay.

Fractions 11 to

19 contained increasing amounts of H particles with progressively increasing densities, and correspondingly decreasing amounts of L particles. After polymerase reaction the products in all cases were P particles; only the H particles had reacted and the L particles were unchanged.

The

largest increase in density (and consequently the largest amount of RNA synthesis per H particle) was observed in the conversion of the H particles in fraction 11 to P particles.

Analytical caesium chloride density gradient

centrifugation (Fig. 3c) showed that this fraction, before reaction, consisted mainly of Ll particles (inactive), together with small amounts of L2 particles (inactive) and H particles (active), present as an asymmetry on the more dense side of the Ll peak.

The products of RNA polymerase

reaction from this fraction were mainly PI particles (containing RNA-Sl) with a small amount of P2 particles (containing RNA-S2) \ the L particles (about 95% of fraction ll) were unchanged (Fig. 3d). It is clear that since the H particles in this fraction had densities between those of Ll and L2 particles, the single-stranded RNA molecules in these H particles must have molecular weights between 0 and 0.17 x 10 molecular weights of RNA-Sl and RNA-S2).

1896

(the difference in

As expected, progressively smaller

Nucleic Acids Research increases in buoyant densities were observed in the conversion of H particles in fractions from 12 to 19 to P particles. Because fractions 11 to 19 contained decreasing amounts of L particles which were unchanged after polymerase reaction, in order to measure the specific activity of the RNA product, it was necessary to separate the P particles from the unreacted L particles.

P particles were isolated by

preparative isopycnic caesium chloride density gradient centrifugation of selected reacted fractions, and RNA isolated from each fraction was examined by polyacrylami de gel electrophoresis.

The amount of RNA in each

gel was quantitated by scanning at 260 nm and then each gel was 3 fractionated and the gel slices were counted for H-incorporation. Approximately equal incorporation on a weight basis was found for RNA-S1 and RNA-S2, but since resolution was incomplete, for greater accuracy they were treated as a single species. The G + C content of P. stoloniferum virus S double-stranded RNA has been estimated to be 53%. Using this value it was calculated that 1 pinole of H-UTP incorporated was equivalent to 1.367 ng of RNA synthesised.

It was then possible to calculate the % of

newly synthesised RNA in the P particles from each fraction.

Since each P

particle contained a molecule of template double-stranded RNA in addition to the newly synthesised RNA it was possible to calculate the amount of newly synthesised RNA (in daltons) per virus particle. 10

A value of 0.94 x

was taken for the molecular weight of template RNA in fraction 11,

since this fraction contained mainly RNA-S1 and an average value of 1.03 x 10

was taken for the molecular weight of template RNA in fractions 12 to

19 which contained approximately equal amounts of RNA-S1 and RNA-S2. The results are shown in Table 1. Table 1 Amounts of newly synthesised RNA in P particles isolated from RNA polymerase reactions of fractions of P. stoloniferua virus S. Fraction no. ( F i g . 1)

Amount of RNA i n gel (ng)

pool of 3 H-UTP incorporated

Amount of 3 H-RNA (ng)

« of J H-RNA

11

233

12

647

207

283

14

628

858

15

3405 4408

471

644

48.3 43.8 25.2 14.6

16

2578

178

243

9.4

18

2533

108

148

19

1511

82.3

61.8

113

84.5

5-8 4.2

Amount of J H-RNA per virus particle, (daltons x 10 ) 0.91 0.90 0.52 0.30 0.19 0.12 0.09 1897

Nucleic Acids Research DISCUSSION The results show that, in the RNA polymerase assay used, only H particles, i.e. those containing both double-stranded RNA and singlestranded RNA, were active; L particles, containing only double-stranded RNA and M particles, containing only single-stranded RNA, were inactive. The results in Table 1 show that the maximum amount of RNA synthesis occurred in H particles in fraction 11 (Fig. l), in which the molecular weight of the single-stranded RNA component was the lowest i.e. between 0 and 0.17 x 10

(95% of fraction 11 was comprised of inactive L particles).

48% of the RNA in P particles, isolated from this fraction after polymerase reaction, is newly synthesised.

It is clear therefore, that since P

particles contain a molecule of template double-stranded RNA, in addition to newly synthesised double-stranded RNA of the same molecular weight, P particles must contain two molecules of double-stranded RNA, one of which is newly synthesised.

The difference between the measured amount of newly

synthesised RNA per virus particle (0.91 x 10

daltons) and the molecular

weight of Sl-RNA (0.94 x 10 ) is probably not experimentally significant. However if the short single-stranded RNA molecules in the H particles of this fraction act as primers for the synthesis of double-stranded RNA, a value slightly lower than 0.94 x 10 amount of newly synthesised RNA.

daltons would be expected for the

It is clear that, since PI and P2

particles gave rise only to RNA-S1 and RNA-S2 respectively, PI particles contain two molecules of RNA-S1 and P2 particles contain two molecules of RNA-S2.

Whether the newly synthesised RNA is present in P particles as

one molecule of double-stranded RNA or as two complementary single-strands of RNA, which anneal to form double-stranded RNA on extraction, cannot be determined from the present data.

The results show that the reaction comes

to a halt when one new molecule of double-stranded RNA (or two complementary single-strands of RNA) has been synthesised.

This is probably because the

newly synthesised RNA remains within the particles, sterically preventing any further RNA synthesis.

The synthesis of one new molecule of double-

stranded RNA (or its equivalent) by these H particles and the absence of reaction by L particles shows that the reaction does not occur according to the model of Chater & Morgan

i.e. by filling in of single-stranded RNA

tails on pre-existing double-stranded RNA molecules.

Particles containing

two molecules of double-stranded RNA have been isolated from Aspergillus foetidus

; these particles may well have been formed ^n vivo by an RNA

polymerase reaction similar to the ia^ vitro reaction described here. Although the detailed mechanism of RNA synthesis is not known, the

1898

Nucleic Acids Research results presented here, which show that single-stranded RNA, present in H particles before polymerase reaction, is not present in P particles after reaction, indicate that the reaction may consist of extending and completing single-stranded RNA chains, previously initiated in vivo, which are later converted to double-stranded RNA, either within the particles or by annealing of complementary single-strands of RNA on extraction.

The

progressively lower amount of newly synthesised RNA in P particles derived from H particles in sucrose density gradient fractions 12 to 19 (Table l) is consistent with their progressively increasing densities and hence additional RNA already synthesised _in vivo, and supports the view

that

these particles represent intermediates in double-stranded RNA replication. It is clear that the most dense H particles, labelled HI and H2 in

Fig.

3(a), are those in which the reaction has almost been completed in vivo. Only small density changes (Fig. 3a and 3b) and small amounts of RNA synthesis (Table 1, fractions 18 and 19) were observed in their conversion to PI and P2 particles.

It has been pointed out

that fungi replicate

asynchronously andraycoviruses,which replicate in parallel with their host cells will also multiply asynchronously.

Mycovirus preparations may

therefore be expected to contain intermediates of replication in addition to mature virions. The lack of RNA polymerase activity in L and M particles was not due to the absence of the RNA polymerase polypeptide molecules from the capsid 12 of these particles. It has been shown previously that the capsid of P. stoloniferum virus S particles consists of 120 molecules of polypeptide S2 (molecular weight 42,500) and one molecule of polypeptide Si (molecular weight 55,500).

The two double-stranded RNA molecules of this virus are 12 . Since S2 is clearly

sufficient to code only for these two polypeptides

the capsid structural polypeptide it is likely that polypeptide Si has the RNA polymerase activity.

Empty capsids lacking RNA also lack polypeptide

Si.

No difference was found in the electrophoretic mobilities, molecular 12 weights or proportions of the capsid polypeptides in M,L or H particles Since the capsids of H particles contains RNA polymerase molecules, the capsids of M and L particles must do also.

The absence of RNA polymerase

activity in L particles indicates that RNA synthesis cannot be initiated in the in vitro assay system; RNA synthesis occurs in this system only in those particles (H particles) in which RNA synthesis has been initiated ^n vivo. It is probable therefore that initiation of RNA synthesis in L particles requires specific initiation factors, not present in the _iii vitro RNA polymerase assay system.

The absence of RNA synthesis by M particles

1899

Nucleic Acids Research shows that either the RNA polymerase cannot utilise single-stranded RNA as a template and/or that specific initiation factors were required, which were absent from the in vitro assay system. In the case of cytoplasmic polyhedrosis virus

and reovirus

it has been shown that initiation of transcription Jji vitro is coupled to the 5' -terminal methylation of the nascent RNA, in the presence of the methyl group donor,S-adenosyl methionine (SAM), giving rise to the 5'terminal structures m7G(5» )ppp(5') A m pGp. . . and u^Q(5' )ppp(5')GmpCp.. . respectively, where m G is 7-methylguanosine, A m is 2< -O-methyladenosine m and G is 21-O-methylguanosine. SAM enhanced the rate of transcription by cytoplasmic polyhedrosis virus considerably

, but stimulation of

reovirus transcription by SAM was much less marked

. Synthesis of reovirus

double-stranded RNA in sub-viral particles, containing reovirus messenger RNA as template

, would not be expected to be enhanced by SAM since the

complementary, negative RNA strands of reovirus have been shown to contain unmethylated 5«-termini

, and the reaction automatically terminates when

the single-stranded RNA template has been converted to double-stranded RNA

.

In the case of P. stoloniferum virus S inclusion of SAM in the RNA

polymerase reaction mixture had no effect on the distribution of RNA polymerase activity among the different particle types, on the rate of RNA synthesis, or on the final amounts of double-stranded RNA and P particles formed.

It is clear, therefore, that SAM was unable to initiate

RNA synthesis in M or L particles, or to stimulate double-stranded RNA synthesis in H particles. Double-stranded RNA mycoviruses are unusual in that they replicate in parallel with their hosts and are not released from the host cells 19 (except under special conditions ) , being carried from one host generation to another.

Recently a model for the replication of double-

stranded RNA mycoviruses was proposed

in which it was suggested that

these viruses may replicate by doubling, a method of replication quite different to those of other viruses, including the double-stranded RNA viruses of animals and higher plants

, but which may be particularly

suited to viruses which remain always intracellular.

The results of the

present paper, which show that the virus particle RNA polymerase catalyses the synthesis of one new molecule of double-stranded RNA per virus particle, i.e. a doubling of the virus RNA, is consistent with this model, although it does not prove it.

It is clear that the RNA polymerase

of P. stoloniferum virus S is a replicase, which catalyses the synthesis of the two complementary strands of double-stranded RNA in a single virus

1900

Nucleic Acids Research particle.

This is completely different to the reovirus system, in which

the virion RNA polymerase is a transcriptase, catalysing the synthesis of only single-stranded messenger RNA molecules '

; in this case a second

sub-viral particle, containing single-stranded RNA as template, is 2 17 required for the synthesis of virus double-stranded RNA ' . The P. stoloniferum virus S RNA polymerase therefore represents a completely new system for the synthesis of double-stranded RNA.

P. stoloniferum virus S

and other double-stranded RNA mycoviruses are not members of the virus 21 family reoviridae, defined as "non-enveloped spherical viruses with genomes consisting of several pieces of double-stranded RNA encapsidated within a single virus particle", and which includes viruses of animals and higher plants, such as reovirus, cytoplasmic polyhedrosis virus and wound tumour virus. Although the genome of P. stoloniferum virus S consists of two, and that of other double-stranded RNA mycoviruses several, pieces of double-stranded RNA, each molecule of RNA is separately encapsidated ' . The RNA replicase activity of particles of P. stoloniferum virus S represents a further difference between the reoviridae and a doublestranded RNA mycovirus. ACKNOWLEDGMENTS I thank Mr. I. P. Blench for excellent technical assistance, Dr. G. Ratti for useful discussions and Mr. G. T. Banks for pilot plant fermentations.

The work was supported by a grant from the Science Research

Council which is gratefully acknowledged. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 910. 11. 12. 13. 14. 15. 16.

Wood, H. A. (1973)- J. Gen. Virol. 20, 61-85. Zweerink, H. J. (1974). Nature 247, 313-314. Lapierre, H., Astier-Manifacier, S. & Cornuet, P. (l97l). C. R. Acad. Sci., ser. D., 273, 992-994. Nash, C. H., Douthart, R. J., Ellis, L. F., Van Frank, R. M., Burnett, J. P. & Lemke, P. A. (1973). Can. J. Microbiol. 19, 97-103. Ratti, G. & Buck, K. W. (1975). Biochem. Biophys. Res. Commun., in the press. Chater, K. F. & Morgan, D. H. (1974). J. Gen. Virol. 24, 307-317Buck, K. W. & Kempson-Jones, G. F. (1973). J. Gen. Virol. 18, 223-235Bozarth, R. F., Wood, H. A. 8= Mandelbrot, A. (l97l) Virology 45, 5l6523. Ifft, J. B., Voet, D. H. & Vinograd, J. (l96l). J. Phys. Chem. 65, 1138 -1145. Erikson, R. L. & Szybolski, V. (1964). Virology 22, 111-124. Buck, K. W. & Ratti, G. (1975)• J- Gen. Virol. 27, 211-224. Buck, K. V. & Kempson-Jones, G. F. (1974). J. Gen. Virol. 22, 441-445. Furuichi, Y. (1974). Nucleic Acids Res. 1, 809-822. Furuichi, Y. & Miura, K. (1975). Nature 253, 374-375Faust, M. & Millward, S. (1974). Nucleic Acid Res. 1, 1739-1752. Furuichi, Y., Morgan, M., Muthakrishishnan, S. & Shatkin, A. J. (1975)• Proc. Nat. Acad. Sci. U.S.A. 71, 3979-3983-

1901

Nucleic Acids Research 17. Burke, D. C. & Russel, W. C. (1975) in Control Processes in Virus Multiplication,25th Symposium of Society for General Microbiology, pp. 253-293, Cambridge University Press, Cambridge. 18. Miura, K., Watanabe, K., Sugiura, M, & Shatkin, A. J . (1974). Proc. Nat. Acad. Sci. U.S.A. 71, 3979-3983. 19. Lemke, P. A., Nash, C.A. & Pieper, S. W. (1973), J . Gen. Microbiol. 76, 265-275. 20. Buck, K. V. & Ratti, G. (1975)• Biochem. Soc. Trans. 3, 542-544. 21. Fenner, F. (1975) • -J. Gen. Virol. 26, 215.

1902